基于混合 Copula 的 ETF 配对交易策略

摘要

Pairs trading is a well‐acknowledged speculative investment strategy that is widely used in the financial markets .Distance and co‐integrated methods are the most commonly implemented pairs of trading strategies .How‐ever ,these two approaches are only able to describe the linear dependency structure between stocks .A new pairs of trading strategies based on mixture Copula method is produced .Three elementary Archimedean Copula families are used to construct mixture Copulas ,which describe the conditional dependence between assets .In the high frequency exchange traded fund (ETF) market ,it is deemed to catch more dependence informations and generate more trading opportunities .%配对交易是一种非常普遍的投资策略，被广泛应用于金融市场。距离法和协整法在配对交易策略中使用最普遍，然而，这2种方法只能描述股票之间的线性相关结构。本文基于3类基本的阿基米德Copula函数构成的混合Copula ，定量刻画了资产之间的条件相关性，给出了新的配对交易策略，并将其运用于高频ET F市场。实证分析表明，新的策略可以捕获更多相依信息，得到更多的交易机会。